How Agent Readiness is scored (method v5)
- Score = Read + Place + Act, minus what Trust loses (up to 36), plus emerging standards (up to +3), rounded, between 0 and 100.
- Emerging standards are new and unproven: passing one adds a point, failing one costs nothing, and none of them can clear an Agent ready condition: Agent ready needs 70 points before any of them are added. A standard moves into the score once mainstream agents use it.
- Every check is graded on what we observe from outside, the same way for every company, against published standards.
- A check we could not run for a reason of our own is shown as not measured and costs nothing.
- Without the Read gate, Place and Act earn nothing. Without the Place gate, Act earns nothing.
- A band that rises or drops shows only after two runs agree.
- A score never rewards using our product.
- Your pages: the homepage plus 5 pages drawn once a month across your site's sections. The URLs are published on your Agent Ready page.
Bands
- Not ready yet: under 70, or held below Agent ready by something still missing.
- Agent ready: 70 or more before extra credit, with the Read gate, the Place gate, Organization details complete, a plain-text version, at least one working interface, and no Trust problem holding the score down. A Trust problem holds the score at 69.
Trust: do they see your domain as credible? (up to 36 points lost)
| Check | Points | How it is tested | Standard |
|---|---|---|---|
| Valid certificate | costs 1.5 if it fails | RSA 2048 bits or more, or ECDSA, and no SHA-1 signature. A confirmed failure caps the score at 69. | RFC 5280 (X.509 certificates) |
| Served over HTTPS | costs 1.5 if it fails | The site is served over HTTPS. A confirmed failure caps the score at 69. | CISA BOD 18-01 |
| HTTP redirects to HTTPS | costs 0.75 if it fails | http:// redirects to https://. A confirmed failure caps the score at 69. | CISA BOD 18-01 |
| No downgrade to HTTP | costs 0.75 if it fails | No redirect goes from HTTPS back to HTTP. A confirmed failure caps the score at 69. Part of https-redirect. | CISA BOD 18-01 |
| HSTS | costs 1.5 if it fails | A Strict-Transport-Security header with a max-age of at least one year. Six to twelve months is a warning (0.75); shorter or missing costs 1.5 and never caps. | RFC 6797 (HTTP Strict Transport Security) |
| DNSSEC | costs 1.5 if it fails | DNSSEC is signed and validates. Unsigned costs 1; broken caps. A confirmed failure caps the score at 69. | RFC 4033 (DNS Security Extensions) |
| Mail provider | costs 1.5 if it fails | Mail servers are published, or a null MX (RFC 7505) says there is none. Neither costs 1.5 and never caps. | RFC 7505 (A Null MX) |
| SPF | costs 0.75 if it fails | A valid SPF record. No record costs 1.5 and caps; ?all or no all costs 1.5 and does not cap; +all, two records or over 10 lookups cost 1.5 and cap. SPF never costs more than 1.5. A confirmed failure caps the score at 69. | RFC 7208 (Sender Policy Framework) |
| SPF within 10 lookups | costs 0.75 if it fails | SPF stays within 10 DNS lookups; over 10 costs 1.5 and caps. Does not apply when the SPF record is missing or already fails. A confirmed failure caps the score at 69. Part of spf. | RFC 7208 (Sender Policy Framework) |
| DMARC | costs 0.5 if it fails | A DMARC record with p=reject or p=quarantine at pct=100. pct under 100 costs 1 and does not cap; p=none costs 1.5 and caps; no record costs 3 and caps. Those are the whole cost: the subdomain and reporting checks do not apply then. A confirmed failure caps the score at 69. | RFC 7489 (DMARC) |
| DMARC for subdomains | costs 0.5 if it fails | DMARC covers subdomains. Does not apply when DMARC is p=none or missing. Part of dmarc. | RFC 7489 (DMARC) |
| DMARC reporting | costs 0.5 if it fails | DMARC reports go to an address, and an external one has authorized it. Does not apply when DMARC is p=none or missing. Part of dmarc. | RFC 7489 (DMARC) |
| DKIM | costs 1.5 if it fails | A DKIM key under a selector we can find; caps only when confirmed broken. A confirmed failure caps the score at 69. | RFC 6376 (DomainKeys Identified Mail) |
| No subdomain takeover | costs 1.5 if it fails | One GET of each known CNAME'd subdomain; a finding is a takeover-prone service showing its own unclaimed-site page. A confirmed failure caps the score at 69. | OWASP Subdomain Takeover |
| Certificate not near expiry | costs 1 if it fails | The certificate is not near expiry. | RFC 5280 (X.509 certificates) |
| CAA | costs 0.5 if it fails | A CAA record. | RFC 8659 (DNS CAA) |
| CAA names an issuer | costs 0.5 if it fails | The CAA record names an issuer. Never caps. Part of caa. | RFC 8659 (DNS CAA) |
| MTA-STS | costs 0.5 if it fails | An MTA-STS policy. | RFC 8461 (SMTP MTA Strict Transport Security) |
| TLS reporting | costs 0.5 if it fails | A TLS reporting address. Part of mta-sts. | RFC 8460 (SMTP TLS Reporting) |
| Old TLS versions turned off | costs 1 if it fails | TLS 1.0 and 1.1 are turned off. | RFC 8996 (Deprecating TLS 1.0 and 1.1) |
| No insecure content on the homepage | costs 1 if it fails | No insecure content on the homepage, and no form that posts to http://. | MDN: Mixed content |
| Domain not near expiry | costs 1 if it fails | The domain registration is not near expiry. | Stringer method A lapsing registration takes the site down; only expiry is observable to agents. |
| Bare domain and www both work | costs 1 if it fails | The bare domain and www both work. | Stringer method Agents do not reliably add or strip www. |
| Content-Security-Policy | costs 1 if it fails | A Content-Security-Policy header. | MDN: Content-Security-Policy |
| Frame protection | costs 1 if it fails | Frame protection (X-Frame-Options or frame-ancestors). | OWASP Clickjacking Defense |
| Privacy and terms pages | costs 1 if it fails | The homepage links a privacy page and a terms page. | Stringer method A security review asks where the privacy policy and terms are. |
| No known-vulnerable scripts | costs 1 if it fails | JavaScript libraries matched by retire.js; caps when a script loads from a known-compromised host. Server and CMS versions are information only. A confirmed failure caps the score at 69. | OWASP A06 Vulnerable and Outdated Components |
| Consent banner | costs 1 if it fails | A consent banner, graded by banner type. | ePrivacy Directive Art. 5(3) |
| No tags before consent | costs 1 if it fails | No tags fire before consent. | ePrivacy Directive Art. 5(3) |
| No analytics before consent | costs 1 if it fails | No analytics fire before consent. | ePrivacy Directive Art. 5(3) |
| Global Privacy Control honored | costs 1 if it fails | A Global Privacy Control signal is honored. | Global Privacy Control |
| Opt-out honored | costs 1 if it fails | Tags stop after an opt-out. Does not apply when there is no banner. | Stringer method After a reject on the consent banner, tags and analytics do not run. |
| X-Content-Type-Options | costs 0.5 if it fails | X-Content-Type-Options: nosniff. | MDN: X-Content-Type-Options |
| Referrer-Policy | costs 0.5 if it fails | A Referrer-Policy header. | MDN: Referrer-Policy |
| Permissions-Policy | costs 0.5 if it fails | A Permissions-Policy header. | MDN: Permissions-Policy |
| security.txt | costs 0.5 if it fails | A security.txt (RFC 9116) that has not expired. | RFC 9116 (security.txt) |
| Universal Analytics retired | costs 0.5 if it fails | No leftover Universal Analytics tag. | Stringer method Universal Analytics stopped processing in 2023; a leftover tag is dead weight. |
| Secure cookies | costs 0.5 if it fails | Cookies the homepage sets carry the Secure attribute. | RFC 6265 (HTTP State Management) |
Privacy and consent checks count in Trust. They take up to 6.5 of Trust's 36 points and never hold a website below Agent ready on their own. This shows what our checker observed on your website. It is not legal advice and not a compliance certification.
Read: can agents access and read your site? (35 points)
| Check | Points | How it is tested | Standard |
|---|---|---|---|
| Agents can open your homepage | 4 | Two looks from our own servers: our signed checker, and a plain browser. Full points when our checker reads the homepage; half when only the plain browser does; none when neither does. Reading means no block or challenge and readable main text in the raw HTML. | Stringer method Two looks from our own servers: a signed checker and a plain browser. |
| AI crawlers allowed | 5 | robots.txt (RFC 9309) lets in the agents that act for a user and the answer engines, on the homepage and the sampled paths. Training crawlers are information only. An unreadable robots.txt fails. | RFC 9309 (Robots Exclusion Protocol) |
| Sitemap | 4 | A sitemap we can fetch and parse. | sitemaps.org protocol |
| Plain-text version | 4 | llms-full.txt, or Markdown (Accept: text/markdown, or a fetchable rel=alternate Markdown link) on the homepage and at least half of the sampled pages. The homepage alone is partly. Markdown content negotiation is shown here as evidence, not scored apart. | Stringer method Nothing published covers a plain-text copy of a site; Markdown is requested with the Accept header of RFC 9110. |
| Pages agents can read | 3, averaged over the graded pages | On each fetched page: no block, and readable main text in the raw HTML. Partly when only the plain browser read it. | Stringer method The fetched page has no block and readable main text in raw HTML. |
| Title, description and excerpt | 3, averaged over the graded pages | On each fetched page: a title, a meta description and an og:description, all non-empty. Partly when one or two are. | HTML Standard: title and meta |
| Headings | 2, averaged over the graded pages | On each fetched page: at least one H1, plus an H2 when the main text is over 300 words. | HTML Standard: headings |
| Clean copy | 2, averaged over the graded pages | On each fetched page: no unbalanced quotation marks, garbled characters or unfilled template text in the title, description, headings or main text. Code blocks, inch marks and « » are ignored. | Stringer method No unbalanced quotation marks, garbled characters or unfilled template text. |
| Markup an agent can parse | 3, averaged over the graded pages | On each fetched page: JSON-LD in the fetched HTML that parses and is valid schema.org. Markup added only by script does not count. | schema.org |
Place: do they put you in the right category? (35 points)
| Check | Points | How it is tested | Standard |
|---|---|---|---|
| Organization on your homepage | 5 | An Organization node in the homepage JSON-LD. | schema.org |
| Organization details complete | 8 | A description, three or more official profiles (sameAs; a profile counts unless it is confirmed broken), and three of: contact, address, legal name, founding date, related organizations. | schema.org |
| Link preview when shared | 2 | og:title, og:description and og:image on the homepage. | The Open Graph protocol |
| Canonical link | 1 | A canonical URL on the homepage. | RFC 6596 (The Canonical Link Relation) |
| Says what it offers and to whom | 5, averaged over the graded pages | On each fetched page: a category and a specific audience ("Everyone" or "All" does not count). On the homepage, knowsAbout and areaServed count instead. | schema.org Audience |
| Complete for its type | 7, averaged over the graded pages | On each fetched page: every main node meets the list for its type. | schema.org |
| Agrees on who you are | 7, averaged over the graded pages | On each fetched page: the publisher or author organization carries the homepage Organization's name, URL and logo, and links to it. | schema.org publisher |
Act: can agents act with you? (30 points)
| Check | Points | How it is tested | Standard |
|---|---|---|---|
| WebMCP | counts toward the interface ladder | A read-only WebMCP tool on the homepage or the contact, demo or pricing page, called by our browser, answers with content. | WebMCP draft (W3C Web Machine Learning Community Group) |
| MCP server | counts toward the interface ladder | initialize, then resources/list and resources/read. Otherwise one tool, only if it is marked read-only, its name reads as a read, and it needs no arguments. We never send arguments that could write. | Model Context Protocol |
| A2A | counts toward the interface ladder | An agent card names an agent, and asked for a made-up task the agent answers that it has no such task, or with a sign-in challenge an agent can follow. We never send a message. | A2A protocol |
| API agents can use | counts toward the interface ladder | A documented GET, HEAD or OPTIONS with no body returns real content, or a 401/403 with a challenge an agent can follow. Redirects stay on your domains. | OpenAPI Specification |
| Agent checkout | counts toward the interface ladder | ACP or UCP discovery on your domain, and its capabilities read. We never create a cart or send a payment token. | Agentic Commerce Protocol |
| Contact path for agents | 5 | An agent can get from the homepage to the demo, contact or pricing page, and its form or booking link has labeled fields and no visible challenge. Standard embeds and invisible CAPTCHA pass. Never submitted. | Stringer method An agent can get to a labeled, challenge-free form or booking link; never submitted. |
| OAuth and OpenID discovery | 1 | Required when we see a real sign-in: a 401 or 403 sign-in challenge on one of our proof calls, OAuth protected-resource metadata, or OAuth or OpenID metadata we found. Then: sign-in server metadata (RFC 8414 or OpenID discovery) at your site's root or through your protected resource metadata, with an issuer that matches where it is published. A WebMCP tool, an open MCP server or API, or an interface we found but could not try is no sign-in: then it does not apply. | RFC 8414 (OAuth 2.0 Authorization Server Metadata) |
Emerging standards: optional, for getting ahead (up to +3)
| Check | Points | How it is tested | Standard |
|---|---|---|---|
| STARTTLS on mail servers | +1 when it passes; never a cost | The mail servers offer STARTTLS. Extra credit, and not measured until a probe host can reach port 25; then Fixes offers a prompt to check it yourself. | RFC 3207 (SMTP over TLS) |
| Verified bot identity | +1 when it passes; never a cost | Our checker's identity can be verified by the site (Web Bot Auth). | Web Bot Auth |
| llms.txt | +1 when it passes; never a cost | An llms.txt that is served and parses. | llms.txt proposal |
| AI use preferences | +1 when it passes; never a cost | A line in robots.txt that says how your content may be used for search, AI answers and AI training, parsed from a `Content-Signal:` directive. | Cloudflare's robots.txt rules for AI use |
| Agents can find your AI tools | +1 when it passes; never a cost | A discovery file (agents.json, A2A card, MCP server card, skills index, resource catalog, API catalog, ARD manifest, or an llms.txt link) points at an interface that passes its proof call. | Stringer method A discovery file counts only when it points at an interface that passes its proof call. |
| Link headers | +1 when it passes; never a cost | Extra credit. Your homepage sends Link headers that point agents at an API catalog or a machine-readable description. Never a failure. | RFC 8288 (Web Linking) |
| DNS-AID | +1 when it passes; never a cost | Extra credit. Your DNS publishes an SVCB agent discovery record under _agents. Never a failure. | DNS for AI Discovery (draft-mozleywilliams-dnsop-dnsaid) |
| auth.md agent sign-up | +1 when it passes; never a cost | Extra credit. Your sign-in server's metadata carries an auth.md agent_auth block with the fields auth.md requires. Never a failure. | auth.md |
| Personal Agent Protocol | +1 when it passes; never a cost | Listed with the extra credit, worth 0 points until the Personal Agent Protocol has a published specification. /.well-known/poppy.json with a protocol version that names your domain. Never a failure. | Stringer method The Personal Agent Protocol has no published specification yet; we read /.well-known/poppy.json and check it names your domain. |
| x402 payments | +1 when it passes; never a cost | Extra credit where we detect agent checkout or a payment interface (the three payment checks earn +1 between them): your site asks for payment the x402 way. Otherwise not applicable, and not shown. | x402 |
| Machine Payments Protocol | +1 when it passes; never a cost | Extra credit where we detect agent checkout or a payment interface (the three payment checks earn +1 between them): your OpenAPI description says which requests take payment, the Machine Payments Protocol way. Otherwise not applicable, and not shown. | Machine Payments Protocol (payment discovery) |
| Agent Payments Protocol (AP2) | +1 when it passes; never a cost | Extra credit where we detect agent checkout or a payment interface (the three payment checks earn +1 between them): your A2A agent card declares the AP2 extension. Otherwise not applicable, and not shown. | Agent Payments Protocol (AP2) |
| BIMI | +1 when it passes; never a cost | Optional: a BIMI record. Never a failure. | BIMI Group |